AutoQoS Configuration

This article covers configuration of AutoQoS VoIP and AutoQoS for the Enterprise. One of the common problems that can arise when designing a network is ensuring that the traffic that is being transported is able to get the appropriate amount of attention and priority when it needs it. For example, if implementing Voice over IP (VoIP) on a network, it is important that this traffic get priority over other types of lower priority traffic. This traffic has to meet a number of traffic requirements in order for the service to work correctly, i.e., the voice is transported without much delay or loss. The way to ensure that this happens is to implement some type (or types) of Quality of Service (QoS). The problem however, that arises many times in smaller companies is that the experience required for good QoS configurations is hard to find. In order to fill this gap, the AutoQoS feature has been developed. When used, the AutoQoS feature is able to automatically configure a device with appropriate QoS commands to ensure correct traffic priority over the network.

There are actually two different AutoQoS feature options, the first is specific to VoIP traffic and requires little configuration in order to implement. The second feature option takes a more detailed approach and is split into two separate configuration phases: a discovery phase and a generation and installation phase.

AutoQoS – VoIP

As is obvious from its name, the AutoQoS – VoIP feature implements a QoS configuration that is specific to VoIP requirements. This feature will determine the correct configuration parameters (Using Classification, Low Latency queueing, compressed Real-Time Protocol (cRTP) and Link Fragmentation and Interleaving (LFI) among other features) based on the type of interface and the configured bandwidth of the interface. As the generation of the configuration using the AutoQoS – VoIP feature happens at the time of command input, the specific configuration generation depends on the correct bandwidth being preconfigured on the interface. Ensure that this is completed before issuing the AutoQoS commands or the configuration will need to be regenerated. It is important to note that the AutoQoS – VoIP feature is only supported on Serial interfaces with PPP or HDLC, Frame Relay DLCI’s (point-to-point), ATM PVC’s and Frame Relay to ATM internetworking links.

AutoQoS – VoIP Configuration

There are a couple of things to keep in mind before issuing the AutoQoS command, these included:

  • Sub-interfaces on serial interfaces are not supported
  • Bandwidth configuration is required before configuration to provide proper configuration generation
  • Interfaces with bandwidths with 768 or below require an IP address configured, this is because on these interfaces Multilink PPP (MLP) is configured.
  • No existing QoS policy can be configured on the interface

The configuration of the AutoQoS – VoIP feature itself is rather simple; the command syntax required are shown below:

router(config-if)#auto qos voip

Once this command is issued, a configuration will be generated and installed. Use the show auto qos {interface interface-type} command to view the generated configuration.

AutoQoS for the Enterprise

As with the AutoQoS – VoIP feature, the AutoQoS for the Enterprise feature is used to automatically configure appropriate QoS configuration. It utilizes the same features to provide a QoS configuration as with the AutoQoS – VoIP feature. However, the AutoQoS for the Enterprise extends on the capabilities offered on the AutoQoS – VoIP feature. Instead of being limited to the generation of a configuration specific to VoIP, this feature provides the ability to monitor the active traffic on the interface and generate a QoS configuration. This discovery is done with Cisco’s Network Based Application Recognition (NBAR) so protocols that are supported by this feature will be cataloged and based on this information, a configuration can be generated specific to the requirements of a specific interface. The time allowed for the device to stay in this discovery phase is important as the feature must be able to get a complete picture of the traffic being sent over the interface; as this differs depending on the situation, the ‘correct’ time is hard to estimate.

AutoQoS for the Enterprise Configuration

As with the AutoQoS – VoIP feature, the same list of things to consider exist with the AutoQoS for the Enterprise feature:

  • Sub-interfaces on serial interfaces are not supported
  • Bandwidth configuration is required before configuration to provide proper configuration generation
  • Interfaces with bandwidths with 768 or below require an IP address configured, this is because on these interfaces Multilink PPP (MLP) is configured.
  • No existing QoS policy can be configured on the interface

The configuration of the AutoQoS for the Enterprise feature includes two separate phases, the second phase relies on the results from the first and must be issued in order.

The command syntax required to start the AutoQoS for the Enterprise discovery phase is shown below:

router(config-if)#auto discovery qos

As stated above, the amount of time to allow the discovery to run depends on the specific network parameters.

Once the discovery has been able to run for its determined amount of time, the second template generation and installation phase can be run; the command syntax required to start this phase is shown below:

router(config-if)#auto qos

Summary

A common problem that exists in many organizations is that the knowledge required to configure QoS correctly can be difficult to obtain. The AutoQoS feature provides the organization with the ability to support QoS with a minimal amount of knowledge, while still obtaining many of the benefits. Hopefully this article has introduced you to the AutoQoS VoIP and AutoQoS for the Enterprise features, and explained how to take advantage of their capabilities.